UvA-DARE (Digital Academic Repository)
E-mental health interventions for harmful alcohol use: research methods and
outcomes
Blankers, M.
Publication date
2011
Link to publication
Citation for published version (APA):
Blankers, M. (2011). E-mental health interventions for harmful alcohol use: research methods
and outcomes.
General rights
It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).
Disclaimer/Complaints regulations
If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible.
Chapter 2
Internet-Based Self-Help for Harmful Alcohol Use:
First Experiences
Chapter based on ůĂŶŬĞƌƐ͕D͕͘<ĞƌƐƐĞŵĂŬĞƌƐ͕Z͕͘^ĐŚƌĂŵĂĚĞ͕D͕͘Θ^ĐŚŝƉƉĞƌƐ͕'͘D͘;ϮϬϬϳͿ͘ ĞƌƐƚĞ ƌǀĂƌŝŶŐĞŶ ŵĞƚ /ŶƚĞƌŶĞƚͲĞůĬƵůƉ ǀŽŽƌ WƌŽďůĞĞŵĚƌŝŶŬĞƌƐ &ŝƌƐƚ džƉĞƌŝĞŶĐĞƐ ǁŝƚŚ /ŶƚĞƌŶĞƚͲĂƐĞĚ ^ĞůĨͲ,ĞůƉ ĨŽƌ WƌŽďůĞŵ ƌŝŶŬĞƌƐ͘ DĂĂŶĚďůĂĚ 'ĞĞƐƚĞůŝũŬĞ ǀŽůŬƐŐĞnjŽŶĚŚĞŝĚ͕ ϲϮ͕ ϭϬϯϮͲ ϭϬϰϯ͘ ůĂŶŬĞƌƐ͕ D͕͘ <ĞƌƐƐĞŵĂŬĞƌƐ͕ Z͕͘ ^ĐŚƌĂŵĂĚĞ͕ D͕͘ EĂďŝƚnj͕ h͕͘ Θ ^ĐŚŝƉƉĞƌƐ͕ '͘ D͘ ;ϮϬϬϴͿ͘ /ŶƚĞƌŶĞƚƉƌŽŐƌĂŵŵ ^ĞůďƐƚŚŝůĨĞ ůŬŽŚŽů͗ ƌƐƚĞ ƌŐĞďŶŝƐƐĞ /ŶƚĞƌŶĞƚ ^ĞůĨͲ,ĞůƉ WƌŽŐƌĂŵ ůĐŽŚŽů͗&ŝƌƐƚ džƉĞƌŝĞŶĐĞƐ͘^ƵĐŚƚ͕ϱϰ͕ϮϳϵͲϮϴϳ͘Fir
st
Experiences
Abstract
KďũĞĐƟǀĞ
To evaluate an internet-based self-help programme for harmful alcohol use (IS) with respect to participant characteristics, intervention uptake, treatment satisfaction, and clinical outcomes.DĞƚŚŽĚƐ
In a retrospective cohort study, data from 3,386 participants who subscribed between September 2003 and July 2005, and survey responses of 290 participants have been analysed.Results
An average 154 participants subscribe to IS each month. Participants are almost equal proportions males and females in their 40s; the majority is highly educated and employed. They consume on average 40 standard units of alcohol per week. Most have never been in formal treatment. Adherence to IS is limited, the majority of subscribers to the intervention visits the IS website only on a few occasions. Almost half of the participants suggests improving the possibilities for interaction. During and after participation in IS, a reduction in alcohol use is observed in comparison to baseline.ŽŶĐůƵƐŝŽŶ
The internet provides new possibilities to address harmful users of alcohol. E-mental health interventions may integrate in mental healthcare treatment systems over the next few years, if they prove to be (cost-)effective.Chapt
er 2
Introduction
The popularity of Internet-based healthcare has been on the rise during the first decade of the 21st century. In 2004, 75% of Internet users in the Netherlands have searched the Internet for information on somatic or mental health care (van Rijen, 2005). In a study on e-mental health developments in the Netherlands, Riper, Smit, van der Zanden and colleagues (2007) report on the availability of internet-based interventions for depression, anxiety, harmful alcohol and substance use and other forms of psychopathology. Riper et al. (2007) conclude that the wide array of available e-mental health interventions can help many people address mental health problems, especially when the threshold to get regular outpatient treatment is too high. The internet is suggested as an option to improve the accessibility of mental health care. It could be crucial to reduce the treatment gap (World Health Organization, 2005) in mental health care. In order to do so, it is a prerequisite that internet-based mental health interventions are easily accessible and show proof of clinical effectiveness.
The Dutch e-mental health overview study by Riper and colleagues (2007) presents the developments over the last 10 years. The study reports on 14 different internet-based interventions for depression, 6 interventions for anxiety disorders and 18 programs for harmful alcohol use. Most of the interventions are based on cognitive behavioural therapy (CBT), motivational interviewing (MI) techniques, or a combination of the two. For only a minority of the interventions, evidence regarding the effectiveness is available, but the evidence base for internet-based interventions is growing. In a meta-analysis, Spek, Cuijpers, Nyklidek and colleagues (2007) combined the outcomes of 12 randomized clinical trials (RCTs) on internet-based interventions for anxiety and depression. The combined effect size for interventions focussing on depression was small (Cohen’s d=0.27), the effect size for anxiety was large (Cohen’s d=0.96). For both internet-based depression and anxiety interventions, therapist support led to larger average effect sizes (Spek et al., 2007).
Little has been published on the effectiveness of internet-based interventions for harmful use of alcohol and other substances, up until the year 2004 (Copeland & Martin, 2004). Recently, an RCT was published on internet-based alcohol self-help (Riper, Kramer, Smit, Conijn, Schippers, & Cuijpers, 2008). 261 participants were randomly assigned to either internet-based self-help or to a control group in which they only received a printed information leaflet. This six-week self-help program is based on CBT and does not involve interaction with a
Fir
st
Experiences
therapist. The most important outcome measure in this study was self-reported alcohol consumption, six months after randomization. At baseline, average daily consumption was about five standard drinking units per day (containing 10 grams of ethanol). After six months, a significant reduction of two glasses per day in the internet-based self-help group versus a reduction of only half a glass per day in the control group was found.
In 2006, all mayor substance abuse treatment centres (SATCs) in the Netherlands have a website with general information about treatment options, entrance criteria and contact details. Some SATCs provide more advanced, interactive websites with animations, video, tests with (algorithmically) tailored feedback, and internet-based interventions. The SATC participating in the studies presented in this dissertation has been developing internet-based self-help interventions for alcohol-, cannabis-, cocaine- and tobacco users and for problematic gamblers since 2003. These self-help interventions are publicly available, free of charge. In this chapter, initial experiences with the SATC’s internet-based self-help intervention for harmful alcohol use (Internet-based Self-help, IS) will be discussed.
IS was developed by a working group consisting of addiction experts, prevention workers and web-developers. Several of the intervention’s exercises are interactive translations of sections from an existing CBT/MI treatment manual (de Wildt, 2000). This treatment manual is widely used by SATCs in the Netherlands. As an advantage over regular outpatient treatment, no costs are involved with participation in IS, and therefore no insurance company or general practitioner will have to be informed on their clients participation. After a successful testing phase, IS is available since September 2003. From this moment on, the working group shifted its focus to optimization and research on the effectiveness of IS. The current study is the first pilot evaluation, which aimed to describe IS’ participants, and explore intervention adherence, treatment satisfaction, and drinking behaviour change. These aims led to the following research questions:
(a) How many participants start using IS on average per month? (b) How can participants be characterized demographically? (c) How well do participants adhere to the intervention? (d) Are participants satisfied with IS?
(e) What suggestions do participants make for improvement? (f) Do participants change their alcohol use after they subscribed?
Chapt
er 2
Method
DĞĂƐƵƌĞƐ
For this retrospective cohort study, all newly subscribed IS participants between the intervention’s launch in September 2003 and July 2005 were invited to participate – if they had provided a valid email address. For this 22 month period, all activities in IS were saved in the database for all participants. For the current study, IS activity data was combined with survey data collection from subscribers who were willing to participate.
The demographic characteristics of the study participants were collected using a subscale of the Addiction Severity Index (Hendriks et al., 1989). Questions on the reasons for participation in IS were developed for the purpose of this study. Questions on participant satisfaction were taken from the Dutch mental healthcare satisfaction scale (GGz thermometer; Kok & Mulder, 2005). This mental healthcare satisfaction scale has often been used in substance abuse treatment evaluation studies in the Netherlands. Survey items to measure alcohol use were taken from the Addiction Severity Index, combined with 7-day timeline follow back methodology (Sobell & Sobell, 1992) to improve the recall of consumed quantities. The amount of alcohol consumed is measured in standard 10 g ethanol drinking units.
dŚĞ/^/ŶƚĞƌǀĞŶƟŽŶ
The primary aim of IS is to offer the participants insight in their drinking behaviour and to assist them in their attempt to change this behaviour. Participants are working on the treatment modules and exercises by themselves, without assistance of a therapist. The self-help modules are derived from an outpatient CBT/MI protocol. According to this implementation of CBT/MI, alcohol- or drug use is context-dependent, acquired behaviour. The intervention exists of learning to recognize and avoid contextual factors triggering consumption, contemplating ways to regulate emotions or urges, and the training of skills to resist craving (Waldron & Kaminer, 2004). A central element is monitoring and reporting of alcohol consumption by the participants. At times of the subscription, each participant is asked about his or her alcohol use during the preceding week. After subscribing, the participant can monitor alcohol use, craving and relevant situational cues on a daily basis using the different treatment modules available in the program. Both the subscription and participation procedures are highly automated; the number of people that can participate simultaneously in IS is
Fir
st
Experiences
therefore not dependent on availability of SATC staff. During the subscription procedure, participants log in to the module and choose their own username and password. If they also provide an email address, they can subscribe to an automated email reminder. If they do, they will receive an email from IS after two weeks of inactivity.
Figure 2.1 ĞƐŝŐŶŽĨƚŚĞ/^/ŶƚĞƌǀĞŶƟŽŶ
Figure 2.1 presents the different modules and exercises of IS. During the subscription procedure, the subscriber is first asked to report his or her alcohol use in the last week. Next, the advantages and disadvantages of drinking and moderation are reviewed. Then a personal goal is set: to moderate alcohol use, or to quit drinking alcohol. After the personal drinking goals are set, attention is focused on how this goal can be achieved. To conclude the subscription procedure, a reading exercise is provided on how to cope with alcohol craving. Only after subscription is completed, subscribers become participants and are
ZĞŐŝƐƚĞƌƵƐĞƌŶĂŵĞ ůĐŽŚŽů ĐŽŶƐƵŵƉƚŝŽŶ ůĂƐƚǁĞĞŬ ĚǀĂŶƚĂŐĞƐͬ ĚŝƐĂĚǀĂŶƚĂŐĞƐ 'ŽĂůƐĞƚƚŝŶŐ ŽƉŝŶŐ ƐƚƌĂƚĞŐŝĞƐ ůĐŽŚŽů ƵƐĞ ƌĞŐŝƐƚĂƚŝŽŶ ĚǀĂŶƚĂŐĞƐͬĚŝƐĂĚǀĂŶƚĂŐĞƐ ŽƉŝŶŐ ƐƚƌĂƚĞŐŝĞƐ WƌĞǀĞŶƚŝŽŶ ƉůĂŶ ŵĞƌŐĂŶĐLJ ƉůĂŶ ŝĂƌLJ ZĞĂĚŝŶŐĞdžĞƌĐŝƐĞƐ &ŽƌƵŵ &ĞĞĚďĂĐŬŽĨƌĞƐƵůƚƐ ZĞĂĚŝŶŐĞdžĞƌĐŝƐĞ ^ƵďƐĐƌŝƉƚŝŽŶ WĂƌƚŝĐŝƉĂƚŝŽŶ ŶƚĞƌ /^ ZĞƚƵƌŶŝŶŐ ƉĂƌƚŝĐŝƉĂŶƚƐ EĞǁƐƵďƐĐƌŝďĞƌƐ džŝƚ /^
Chapt
er 2
motivated to enter participant area. Here, the six main treatment modules, four reading exercises and a result feedback page are available. A forum provides opportunities for peer-support or for reading previous posts of fellow IS participants. Participants are suggested to work with the IS intervention for a duration of 4 to 6 weeks.
The IS intervention is sustained by four piers. The first pier is reporting alcohol use, goal setting, and identifying risk situations which may lead to relapse. By monitoring alcohol use and reporting the associated contextual cue’s, emotions, and cognitions, the participant learns to recognize risky situations and to make these explicit. A second pier is receiving feedback. This feedback is generated using information reported under the first pier, and is compared with the personal goal. Receiving this feedback helps gaining insight in the process of goal striving and in the fluctuations in alcohol use. The third pier is acquiring skills and knowledge regarding coping with craving, peer pressure and motivation, and assessing personal risk situations. Participants also learn to anticipate on these personal risk situations and learn to cope with (re)lapse. To reflect on acquired knowledge and skills, participants are stimulated to write in their IS diary. The forth pier consist of interaction with fellow participants through the forum facility. On this forum, participants can discuss their experiences with moderation or quitting alcohol use, risk situations, (re)lapse, or the IS intervention in general. This internet-based self-help group is created to facilitate participants to motivate and consult each other.
ZĞƐƉŽŶƐĞŶĂůLJƐŝƐ
Of the 3,386 IS participants, 1,975 had provided an email address. An invitation to participate in this study was sent to the available 1,975 email addresses. 290 Participants participated and filled out the survey, a 15% response rate. A comparison between responders to the survey invitation (n=290) and all other IS participants (n=3,096) is possible on a limited number of variables. Based on this comparison, it is concluded that responders and all other IS participants do not differ with regard to sex-distribution, level of education and baseline alcohol consumption. Differences between responders and all other IS participants have been found with regard to age (responders M = 43.7 years, other IS participants M=40.0 years, t(3230)=6.02, p<0.05) and duration of IS participation (responders M=25.0 days, other IS participants M=8.9 days, t(3218)=3.22, p<0.05). Based on the IS database and the survey responses, the participants have been described.
Fir
st
Experiences
Results
WĂƌƟĐŝƉĂŶƚƐ
With a total number of participants of n=3,386 over 22 months, an average of 154 participants subscribes to IS per month. These participants are 40 years old on average. 41% is female; 61% is highly educated (Bachelor degree or higher) according to the International Standard Classification of Education (United Nations Educational, Scientific and Cultural Organization, 1997). 27% lives in Amsterdam, the Netherlands, or in one of its suburbs. Amsterdam is the traditional service area of the SATC that developed IS. 68% is employed, either full time or part time. In the week before subscribing to IS, participants consumed an average of 40 standard drinking units.
Based on the survey data, we can further describe the IS participants. 45% considers their alcohol use pattern as ‘risky’; 21% considers him- or herself as severely dependent on alcohol. For 21% of the respondents, IS was the first intervention they have tried in order to change their alcohol use pattern. 18% had received regular treatment from a SATC before subscribing to IS. 22% has discussed their alcohol use with their general practitioner. 39% reported to have participated in informal care, ranging from Alcoholic Anonymous meetings to participation in internet discussion forums.
Figure 2.2 Adherence to IS EŽƚĞ͘ ŝƐƚƌŝďƵƟŽŶ ŽĨ ϯ͕ϯϴϲ ƉĂƌƟĐŝƉĂŶƚƐ ƌĞŐĂƌĚŝŶŐ ƚŚĞŝƌ ĂĚŚĞƌĞŶĐĞ ƚŽ /^͗ ƉŽŽƌ ĂĚŚĞƌĞŶĐĞ͕ Ϯ͕ϬϮϭ ƉĂƌƟĐŝƉĂŶƚƐ ;ϲϬйͿ͖ ŵŽĚĞƌĂƚĞ ĂĚŚĞƌĞŶĐĞ͕ ϵϵϭ ƉĂƌƟĐŝƉĂŶƚƐ ;ϮϵйͿ ĂŶĚ ŽƉƟŵĂů ĂĚŚĞƌĞŶĐĞ͕ ϭϰϮ ƉĂƌƟĐŝƉĂŶƚƐ;ϰйͿ͘ĚŚĞƌĞŶĐĞĚƵƌĂƟŽŶŝƐƵŶŬŶŽǁŶĨŽƌϮϯϮ;ϳйͿŽĨƚŚĞƉĂƌƟĐŝƉĂŶƚƐ͘ ĚŚĞƌĞŶĐĞ ĚƵƌĂƚŝŽŶƵŶŬŶŽǁ Ŷ;Ŷ сϮϯϮͿ WŽŽƌĂĚŚĞƌĞŶĐĞ ϭĚĂLJ ϲĞdžĞƌĐŝƐĞƐ Ŷ сϮ͕ϬϮϭ DŽĚĞƌĂƚĞĂĚŚĞƌĞŶĐĞ ϭϴĚĂLJƐ ϭϰĞdžĞƌĐŝƐĞƐ Ŷ сϵϵϭ KƉƚŝŵĂůĂĚŚĞƌĞŶĐĞ ϭϬϯĚĂLJƐ ϯϱĞdžĞƌĐŝƐĞƐ Ŷ сϭϰϮ
Chapt
er 2
ĚŚĞƌĞŶĐĞ
Based on their adherence to IS, participants can be divided in three groups (Figure 2.2). About 60% (n=2,021) of the IS participants has a poor adherence, and visits the intervention only during the subscription day. During this day, they participate on average in 6 modules / exercises. A second group of 29% (n=991) has visited the intervention on more than one day, but did not adhere to the intervention for the suggested 4 to 6 weeks. This group was moderately active, participated on average for 18 days, and worked on 14 of IS’ modules. The final group, with optimal adherence, did participate for at least 4 weeks. This group contains only 4% of all IS participants. This optimal adherence group did participate in 35 modules on average (most can be done multiple times).
dƌĞĂƚŵĞŶƚ^ĂƟƐĨĂĐƟŽŶ
83% of all responders is satisfied with the self-help module interface. The average rating on a 0-10 scale is 6.4. Main reason for responders to use IS was the ability to change their behaviour without giving up perceived anonymity. IS’ flexibility is positively evaluated as well: no obligations and always available. According to the login data, 40% of all activity in the self-help module takes place outside office hours and 27% during the weekends: assistance from a therapist would usually not have been available during these hours. Only a minority of the participants considers outpatient treatment as an acceptable alternative: 20% would have sought regular care if IS was unavailable and 15% will consider doing so in the future if necessary.
Table 2.1 ^ƵŐŐĞƐƟŽŶƐĨŽƌ/ŵƉƌŽǀĞŵĞŶƚŽĨ/^
^ƵŐŐĞƐƟŽŶƐĨŽƌ/ŵƉƌŽǀĞŵĞŶƚ
Table 2.1 provides an overview of the suggestions for improvement. Most often suggested was to improve the possibility of interaction: either with a
/ŵƉƌŽǀĞ Ŷ;йͿ /ŶƚĞƌĂĐƟŽŶǁŝƚŚƚŚĞƌĂƉŝƐƚ ϯϮ;ϮϮйͿ tĞďƐŝƚĞĚĞƐŝŐŶ Ϯϲ;ϭϴйͿ /ŶƚĞƌĂĐƟŽŶǁŝƚŚƉĂƌƟĐŝƉĂŶƚƐ ϭϴ;ϭϮйͿ /ŶƚĞƌĂĐƟŽŶŝŶŐĞŶĞƌĂů ϭϰ;ϭϬйͿ /ŶĨŽƌŵĂƟŽŶ ϳ;ϱйͿ ^ŽŵĞƚŚŝŶŐĞůƐĞ ϮϮ;ϭϱйͿ EŽƚŚŝŶŐ Ϯϳ;ϭϴйͿ
Fir
st
Experiences
therapist, with other participants, or in general. Responders considered the current possibilities for communication insufficient – they would rather have had direct and personal feedback on their exercises in IS, or a personalized email after a period of inactivity. 12% would like to see the possibilities for interaction with other participants improved. The forum was a step in the right direction, but the addition of a chat facility would be a well-received next step. It would improve the possibilities for peer-support and interaction.
Figure 2.3 ůĐŽŚŽůŽŶƐƵŵƉƟŽŶĂƚĂƐĞůŝŶĞ͕ƵƌŝŶŐ/^ĂŶĚĂƚ&ŽůůŽǁͲhƉ
EŽƚĞ͘ŽdžƉůŽƚƐƉƌĞƐĞŶƚƚŚĞĂŵŽƵŶƚŽĨĂůĐŽŚŽůĐŽŶƐƵŵƉƟŽŶĂƚďĂƐĞůŝŶĞ͕ĚƵƌŝŶŐ/^͕ĂŶĚĂƚĨŽůůŽǁͲƵƉ ĨŽƌŶсϭϬϲƉĂƌƟĐŝƉĂŶƚƐ͘
ŚĂŶŐĞƐŝŶůĐŽŚŽůhƐĞ
Do participants change their alcohol use while participating in IS? Figure 2.3 presents box plots of the self-reported alcohol use of 106 participants that reported their alcohol use at baseline (in the subscription phase), during IS and in the follow-up survey. Before participation, 50% consumed between 4 and 8 units per day (Median=5.5 units). During IS, 50% reduced their consumption to
Ϭ Ϯ ϰ ϲ ϴ ϭϬ ϭϮ ϭϰ ĂƐĞůŝŶĞ ƵƌŝŶŐ/^ &ŽůůŽǁͲƵƉ ^ƚĂŶĚĂƌĚĚƌŝ ŶŬŝ Ŷ ŐƵŶŝƚƐ ĐŽŶƐ Ƶ ŵ ĞĚ ƉĞ ƌĚĂLJ
Chapt
er 2
between 1 and 4 units per day (Median=2.1 units). In the follow-up survey, 50% consumed between 2 and 6 units per day (Median=3.9 units). The differences in units consumed are significant between baseline and during IS, t(105)=8.97, p<0.001; and between baseline and follow-up, t(105)=4.69, p<0.001.
Discussion
The results of this evaluation indicate that an average of 154 participants subscribes to IS per month. Participants are almost equal proportions males and females in their 40s; the majority is highly educated and employed, and drinks on average 40 alcoholic beverages per week. They often consider themselves as risky drinker, however one in five perceives him- or herself as severely dependent on alcohol. Most have never been in formal SATC treatment before IS, and most will not consider to get formal treatment in the future. Adherence to IS is limited, the majority of the participants visits the intervention website only a few times. The majority of the participants are satisfied with IS, but at the same time almost half of them suggests to improve the possibility of interaction, preferably with a therapist. The technical design of the intervention could also be improved. During participation in IS, a reduction in alcohol use is observed. This reduction appears to persist after participation is discontinued.
The internet provides SATCs with new ways to address its target audience. If proven to be effective, e-mental health interventions may integrate in mental healthcare treatment systems over the next few years, as these interventions are able to serve a currently unaddressed target population. The notion that online treatment is able to address previously unseen clients is previously observed by e.g. Postel, de Jong, & de Haan (2005). IS is used by harmful alcohol users for changing problematic and unhealthy drinking behaviour. It is difficult for participants to stay motivated when personalized feedback is lacking, and therefore dropout is common in e-health interventions (Eysenbach, 2005). This apparent lack of personalization and tailored feedback should however not be regarded as inherent to internet-based interventions. Instead, it should encourage to develop and experiment with other forms of guided self-help and internet-based therapy (Schippers, 2007). It urges e-mental health developers to improve the personalized feedback component.
Suggestions regarding forum facilities, design of the intervention, information have been addressed in the year 2006 update of the IS intervention. The suggestion to add therapist support is incorporated in an internet-based
Fir
st
Experiences
therapy (IT) application. In this IT intervention, one-on-one text-based chat sessions between a client and a therapist intend to improve the personalized tailoring and guidance of IS. The updated version of IS and the newly developed IT intervention will be subjected to an RCT in order to conclude on their (cost-) effectiveness. The design and outcomes of this study will be presented in the upcoming chapters of this dissertation.
>ŝŵŝƚĂƟŽŶƐ
As in most naturalistic evaluation studies, this study has a number of methodological weaknesses. First, the survey response-rate of 15% is very low. It is therefore disputable whether the results found and presented in this paper are representative to the full population of IS participants. An extensive non-response analysis, coupled with more effortful measurement methodology, could improve the representativeness of future studies. Second, identifying individual participants based solely on email addresses makes the perceived anonymity high, but is far from ideal from a methodological perspective. It is possible that one single person subscribed more than once to the IS intervention using multiple email addresses, or that more than one person accessed a single treatment environment using a shared email address. Third, the current retrospective design without a control group makes it impossible to conclude on the effectiveness of the IS intervention. Although the observed drinking behaviour changes are in the desired direction, it is impossible to attribute these observed changes to the intervention. The planned RCT, in which we will compare IS and IT will provide the data needed to conclude on the effectiveness of these internet-based interventions for reducing harmful alcohol use.
